Studies with Isolated Hepatocytes and T-Butyl Hydroperoxide

Proc. Nati Acad. Sci. USA Vol. 79, pp. 6842-6846, November 1982 Biochemistry

Regulation of intracellular calcium compartmentation: Studies with isolated hepatocytes and t-butyl hydroperoxide (mitochondrial Ca2" /extramitochondrial Ca2" /NADPH oxidation/thiol homeostasis/arsenazo III) GIORGIO BELLOMO*, SARAH A. JEWELL, HjoRDIs THOR, AND STEN ORRENIUSt Department of Forensic Medicine, Karolinska Institutet, S-104 01 Stockholm, Sweden Communicated by Peter Reichard, August 6, 1982 ABSTRACT In suspensions ofisolated hepatocytes, two intra- dria play a major role in the control of calcium compartmen- cellular Ca2+ pools were distinguished in the presence of the me- tation and the regulation of cytosolic Ca2". It has been dem- tallochrome indicator arsenazo HI, first by treatment with the onstrated by Lehninger et aL (14), and subsequently confirmed uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone by others (15-17), that the pyridine nucleotide redox state is (FCCP) and then with the Ca2+ ionophore A23187. The available a determining factor in the ability of mitochondria to take up evidence indicates that the two pools are of mitochondrial and and retain Ca2+. In addition, Sies et aL (18) recently have shown extramitochondrial origin. Metabolism of t-butyl hydroperoxide that the oxidation of cytosolic pyridine nucleotides can cause by hepatocytes caused release of Ca2+ from both compartments a release of Ca2+ from isolated perfused liver, although the concomitant with oxidation of cellular glutathione and NADPH, which was followed by characteristic alterations in cell surface site(s) from which Ca2+ was released could not be identified. structure. When NADPH oxidation was prevented by selective In the present work we introduce a nondisruptive method inactivation of glutathione reductase, t-butyl hydroperoxide me- for measuring two distinct intracellular Ca2+ pools-mitochon- tabolism was without effect on the mitochondrial Ca2+ pool, drial and extramitochondrial. Using this system, we have ob- whereas the loss from the extramitochondrial pool was acceler- served that significant alterations in hepatocyte Ca2+ homeo- ated. Our results suggest that different regulatory mechanisms stasis are caused by t-butyl hydroperoxide metabolism. Hydro- modulate mitochondrial (NADPH-dependent) and extramitochon- peroxides, known to cause oxidative stress and toxicity in drial (thiol-dependent) Ca2+ compartmentation and that distur- various tissues (cf. 19, 20), are metabolized in hepatocytes by bance ofnormal Ca2+ homeostasis may be critical in peroxide-in- the glutathione peroxidase system (15), leading to glutathione duced cytotoxicity. (GSH) and NADPH oxidation. The metabolism of t-butyl hy- droperoxide has been shown to impair the ability of liver mi- The calcium ion exerts a profound influence on a wide variety tochondria to retain Ca2+ (15-17) and to cause Ca2+ release from ofcellular processes (1-3). The distribution ofCa2" within the perfused liver(18). Our results indicate thatdifferent regulatory cell is complex, involving the binding to cellular components mechanisms control the mitochondrial and extramitochondrial (4, 5) as well as the action of specific compartmentation pro- compartments and that the metabolism of t-butyl hydroperox- cesses (6). The cytosolic Ca2" concentration is maintained at ide causes Ca2+ loss from both intracellular pools. Depletion levels 3-4 orders of magnitude lower than in the extracellular of cell Ca2" is associated with characteristic alterations in he- medium due to active extrusion of the ion by specific Ca2" patocyte surface structure, which may be an early indication of pumps present in the plasma membrane as well as the inter- cytotoxicity. dependent action of intracellular sequestration sites (7). How- ever, at the present time little is known about the physiological MATERIALS AND METHODS mechanisms involved in intracellular Ca2+ homeostasis. Several methods have been proposed to study Ca2" fluxes Materials. Collagenase (grade II) and FCCP (carbonyl cya- from one cellular compartment to another. Claret-Berthon et nide p-trifluoromethoxyphenylhydrazone) were obtained from al. (8), using 'Ca2" loading in the perfused rat liver, proposed Boehringer Mannheim. Arsenazo III, bovine serum albumin a model with three different calcium pools: a dynamic cyto- (fraction V), Hepes, NADH, and ruthenium red were pur- plasmic pool (which can itselfbe subdivided into mitochondrial, chased from Sigma. The cation ionophore A23187 was pur- microsomal, and cytosolic subfractions), a slowly exchangeable chased from Calbiochem-Behring and Percoll was obtained nuclear pool, and an apparently nonexchangeable mitochon- from Pharmacia. BCNU [N,N-bis(2-chloroethyl)-N-nitrosourea] drial pool (insoluble calcium). In isolated cells, compartmen- was agiftfrom Jakob Kaluski ofBristol Laboratories (Stockholm, tation and fluxes of Ca2" have been studied by the use of (i) Sweden). All other reagents were commercial products of the calcium-selective microelectrodes (9); (ii) photoproteins, such highest available grade of purity. as aequorin (10); (iii) metallochromic indicators, notably arsen- Hepatocyte Isolation and Incubation. Hepatocytes were azo III and cell with mea- isolated from male Sprague-Dawley rats (180-200 g; allowed (11); (iv) rapid disruption techniques food and water ad lib) by collagenase perfusion ofthe liver (21). surement of Ca2+ content in the different organelle fractions The yield was 2-4 x 108 cells per liver and, immediately after (12). Recently, Murphy et al. (13) developed a method for the isolation, the hepatocytes excluded both trypan blue and quantitation ofcytosolic free Ca2+ using arsenazo III to measure NADH (90-100%) (cf. 21). Hepatocytes were incubated in Ca2" fluxes through the hepatocyte plasma membrane which Krebs-Henseleit buffer atpH 7.4, supplemented with 12.6 mM was made permeable by digitonin treatment. Data obtained by these methods indicate that the mitochon- Abbreviations: FCCP, carbonylcyanide p-trifluoromethoxyphenylhy- drazone; BCNU, N,N-bis(2-chloroethyl)-N-nitrosourea; GSH, gluta- The publication costs ofthis article were defrayed in part by page charge thione; GSSG, glutathione disulfide. payment. This article must therefore be hereby marked "advertise- * Present address: Clinica Medica IIa, University ofPavia, Pavia, Italy. ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. t To whom reprint requests should be addressed.

6842 Downloaded by guest on September 27, 2021 Biochemistry: BeUomo et al. Proc. Natl. Acad. Sci. USA 79 (1982) 6843 Hepes (22), at a concentration of 6 X 106 cells per ml for Ca2+ afterpreincubation for30 min with orwithout 20 AM ruthenium experiments or 1 X 106 cells per ml for GSH and NADPH as- red. Various concentrations of Ca2+ (up to 5 ,M) were added; says. After a 15-min preincubation period at370C, samples were this was followed by the addition of2 ,uM digitonin. After Ca2+ taken as zero time and t-butyl hydroperoxide was added. Spe- uptake was complete, 10 ,M FCCP was added and subsequent cial pretreatments of the hepatocytes were carried out as fol- release was recorded. lows: (i) To protect the cells from thiol group oxidation during Hepatocyte GSH was measured either by the colorimetric t-butyl hydroperoxide metabolism, hepatocytes were preincu- assay ofSaville (25) or by HPLC as described by Reed et al. (26). bated for 5 min with 2 mM 1,4-dithiothreitol. (ii) To inhibit Cellular pyridine nucleotide levels were determined spectro- mitochondrial Ca2+ uptake (23), samples of hepatocytes were photometrically as described by Klingenberg (27). incubated in the Krebs-Henseleit medium (described above) Scanning Electron Microscopy. Samples were processed by for 30 min at 370C in the presence of either 5 or 20 uM of ru- standard procedures, which involved glutaraldehyde and os- thenium red; this treatment did not affect the viability of the mium fixation followed by critical point drying. A Jeol model hepatocytes. (iii) To inactivate glutathione reductase [NAD- JSM35 scanning electron microscope was used to visualize and (P)H:oxidized-glutathione (GSSG) oxidoreductase, EC 1.6.4.2], photograph the samples. A large number of hepatocytes were hepatocytes first were preincubated in an amino acid-supple- examined and representative cells showing typical morphology mented Krebs-Henseleit medium in the presence of 100 j.M were photographed. BCNU as described by Eklow et al. (24). After 30 min the cells were washed, resuspended, and incubated for an additional 90 min in the same medium, except that BCNU was absent. This RESULTS procedure provides hepatocytes with long-lasting inhibition of A typical pattern of Ca2+ release caused by the addition of glutathione reductase, but with a normal cellular GSH level at FCCP and the ionophore A23187 to isolated hepatocytes is il- the end of the treatment (24). lustrated in Fig. IA. The use ofhigher concentrations ofFCCP Biochemical Assays. For the measurement ofcellular Ca2', did not result in any increased Ca + release over what was ob- the hepatocytes were separated from the Ca2+-containing served with the 10 ,uM concentration. Administration of the Krebs-Henseleit medium by rapid centrifugation through a ionophore A23187 was found to cause complete Ca2+ release suspension of Ca2+- and Mg2--free Hanks' solution (13) and (equal to the total release caused first by FCCP and then by the Percoll (final density: 1.06 g/ml). They then were quickly re- ionophore), and there was no further release upon subsequent suspended in the modified Hanks' medium and were separated addition of FCCP. The FCCP-releasable and FCCP-nonre- into two parts; one part was used for counting the number of leasable cellular Ca2` pools constitute approximately 60% and cells present and assaying cell viability (21), and the other part 40%, respectively, of the total amount of Ca2+ that could be was used for Ca2+ measurement. Centrifugation of the cells released from control hepatocytes incubated under our experi- through the Percoll mixture removes nonviable cells, ensuring mental conditions; this proportion was quite consistent from one that, even at the higher dose of t-butyl hydroperoxide, only cell preparation to another. viable cells were assayed. In all cases but one (see Results) cell A variety of experiments were undertaken to characterize viability remained quite high (>75%) throughout the 30-min these two Ca2+ pools. Administration of the uncoupler dicu- incubation period. marol (30 ,M) produced the same Ca2+ release as FCCP and Intracellular Ca2+ content was determined by dual wave- did not affect the Ca2` release caused by subsequent addition length spectrophotometry (685-675 nm) by using an Aminco of A23187 (not shown). Preincubation of hepatocytes with ru- DW2 UV/VIS spectrophotometer and purified arsenazo III thenium red to block the uptake portion of normal mitochon- (2,2' -[1,8 -dihydroxy-3,6-disulpho -2,7-naphthalene -bis(azo)]- drial Ca2+ cyclingcaused a substantial change in the proportions dibenzenearsonic acid) (11) (final concentration: 30 ,uM). FCCP of the pools; this pretreatment resulted in a dose-dependent (10 ,uM) was added first to the cell suspension (=4 X 106 cells decrease in the size of the FCCP-releasable Ca2` pool as well per ml) and Ca2O release was recorded until no further change as a lesser decrease in the total level of releasable Ca2` (Fig. in absorbance was observed. At this point, the Ca2+ ionophore 1 B and C). Hepatocytes that were preincubated in 20 ,uM ru- A23187 (15 ,uM) was added and Ca2' release was recorded. thenium red and then permeabilized by digitonin treatment In experiments with digitonin-permeabilized hepatocytes, took up none of the Ca2+ added to the extracellular medium the cells were carried through the procedure just described (up to 5 ,M). However, when control cells were permeabilized

c FCCP A FCCP

CO A a A2:

A23187 o0 \ .0

C C-

FIG. 1. Calcium release from hepatocytes by FCCP and by A23187. Cells were incubated and prepared for spectrophotometry as described. At the times shown, 10 ,uM FCCP or 15 AM A23187 was added, and traces typical of those shown were recorded. For the Ca2+/arsenazo III experiments, AA6W_675 = 0.005 per 1 ,uM Ca2 . (A) Control incubation; (B) preincubation of hepatocytes with 5 ,pM ruthenium red; (C) preincubation of he- patocytes with 20 ,uM ruthenium red. Traces are representative of results obtained with five or more separate cell preparations. Downloaded by guest on September 27, 2021 6844 Biochemistry: Bellomo et al. Proc. Nad Acad. Sci. USA 79 (1982)

C A B g A B 100. 0 100 O EloF 0

50- 50-

C D LC D 10 20 30 10 20 30 . b 20 30 lb 20 30 Time (min) Time (min) FIG. 2. Changes induced by the metabolism of t-butyl hydrope- FIG. 3. Effects of 1,4-dithiothreitol on changes induced by the roxide in the levels of cellular GSH (A), NADPH (B), FCCP-releasable metabolism of t-butyl hydroperoxide in the levels of cellular GSH (A), Ca2" (C), and FCCP-nonreleasable Ca2" (D). Cells were incubated as NADPH (B), FCCP-releasable Ca2 (C), andFCCP-nonreleasable Ca2" control (A) or with t-butyl hydroperoxide (o, 1 mM; o, 4 mM). The re- (D). Cells were incubated as control (open symbols) or with 2 mM 1,4- sults are expressed as percentage of the values at time zero and are dithiothreitol (closed symbols) before the addition of t-butyl hydro- typical of three to five trials. The absolute values at this time per 1 peroxide (o,n, 1 mM; o, *, 4 mM) at time zero. The results are ex- X 106 cells were: GSH, 45.2 nmol; NADPH 2.8 nmol; FCCP-releasable pressed as percentage of the values at time zero and represent typical 1.8 nmol; 1.1 values for three to five experiments. The absolute values at this time Ca2", FCCP-nonreleasable Ca2", nmol. were not different from those given in Fig. 2, for both control and 1,4- dithiothreitol-treated hepatocytes. with digitonin in the presence ofextracellular Ca2", rapid Ca2+ uptake ensued; subsequent addition ofFCCP then caused com- prevented) and could be separated from NADPH oxidation plete release of the Ca2+ taken up (not shown). This evidence (Fig. 4 A and B). The responses of the two Ca2" pools to these indicates that the FCCP-releasable Ca2+ may be of mitochon- conditions also could be distinguished; the BCNU pretreatment drial origin. reversed the hydroperoxide-induced loss from the mitochon- In intact hepatocytes, t-butyl hydroperoxide is metabolized drial pool, whereas the extramitochondrial pool became more by the GSH peroxidase (GSH:H202 oxidoreductase, EC rapidly and extensively depleted (Fig. 4 C and D). Inhibition 1. 11. 1.9) system present both in the cytosolic and mitochondrial ofglutathione reductase also enhanced dramatically the toxicity compartments (15). This results in the formation of GSSG, oft-butyl hydroperoxide, because after only 5 min of exposure which is subsequently reduced to GSH by glutathione reduc- to the hydroperoxide the numberofviable cells was insufficient tase at the expense of NADPH. Under our experimental con- to continue Ca2" measurement (see Materials and Methods). ditions, the metabolism of t-butyl hydroperoxide by control Perturbation of the hepatocyte surface structure, with char- hepatocytes led to both a marked decrease in cellular GSH level acteristic blebbing of the plasma membrane, was an early sign (Fig. 2A) and a sharp initial drop in NADPH concentration (Fig. of t-butyl hydroperoxide toxicity (Fig. 5A) occurring within 30 2B). No change in NADH level occurred (not shown). Incu- min of hydroperoxide addition and well before any significant bation of hepatocytes with t-butyl hydroperoxide also affected loss of cell viability. As shown in Fig. 5B, this effect was pre- the size ofboth Ca2+ pools; the size ofthe FCCP-releasable pool vented by the presence of 1,4-dithiothreitol in the incubation was markedly affected only at the higher concentration of the medium, whereas it was markedly enhanced in the BCNU-pre- hydroperoxide (Fig. 2C), whereas the size of the FCCP-non- treated hepatocytes (Fig. 5C). Therefore, this loss of normal releasable Ca2+ pool was affected by both concentrations in a surface morphology appears to be related to the thiol group dose- and time-dependent manner (Fig. 2D). oxidation and depletion of cell Ca2" that results from hydro- To study the relationship between the peroxide-associated peroxide metabolism. oxidation of GSH and NADPH and the decrease in the Ca2+ pools, two series ofexperiments were performed. Cellular thiol groups were protected from oxidation during t-butyl hydro- DISCUSSION peroxide metabolism by the addition of1,4-dithiothreitol to the In this work we have employed a relatively simple method to incubation 5 min prior to addition of the hydroperoxide. This quantitate two different intracellular calcium pools in isolated pretreatment prevented the effects oft-butyl hydroperoxide on hepatocytes. The FCCP-releasable Ca2" pool appears to be GSH and NADPH levels and on the loss ofCa2' from both pools derived from the mitochondrial compartment, measurable as (Fig. 3). In another group of experiments, hepatocytes were a consequence first ofcollapse ofthe mitochondrial membrane pretreated with BCNU, a selective inactivator of glutathione and then a transient release of Ca2" into the cytosol and sub- reductase (28), to minimize the NADPH oxidation that occurs sequent extrusion from the cell. With isolated, coupled mito- during hydroperoxide metabolism. With this treatment GSH chondria, the collapse of the proton electrochemical gradient depletion was accelerated (as reduction ofthe GSSG formed was by uncouplers (such as FCCP and dicumarol) has been shown Downloaded by guest on September 27, 2021 Biochemistry: Bellomo et al. Proc. NatL Acad. Sci. USA 79 (1982) 6845 dicate that the uncoupler's releasing action is specific for the mitochondrial compartment. Moreover, studies done with rat liver microsomes suggest that the endoplasmic reticulum [which is thought to sequester a large majority ofthe cell's non- mitochondrial Ca2+ (13)] is unaffected by uncouplers such as FCCP and dicumarol but is highly sensitive to A23187 (unpub- lished results). Taken together, these facts provide strong sup- port for the mitochondrial origin of the FCCP-releasable Ca2+ pool. The amount of Ca2+ that is releasable from isolated hepato- A B cytes by the cation ionophore A23187 is higher than that re- leasable by FCCP alone and is also higher than the mitochon- ~100 drial Ca2+ content as determined by atomic absorption spec- trometry (13, 29, 31, 32). The ionophore alone releases the same total amount of Ca2+ from hepatocytes as does sequential FCCP/A23187 treatment and also readily induces Ca2+ release 50 0 from isolated microsomes. Therefore, it seems justified to con- clude that the FCCP-releasable and FCCP-nonreleasable Ca2' pools in this study represent Ca2+ derived from the mitochon- drial and extramitochondrial compartments, respectively. IC 0 The observed decrease in the mitochondrial Ca2+ pool during ib0 30 i0b 2b 30 t-butyl hydroperoxide metabolism is in agreement with the re- Time (min) sults previously obtained with isolated mitochondria by our group FIG. 4. Effects of BCNU on changes induced by the metabolism of (17) and others (15). Our results also support the hy- t-butyl hydroperoxide in the levels of cellular GSH (A), NADPH (B), pothesis that the oxidation of mitochondrial pyridine nucleo- FOOP-releasable Ca"~(C), and FOOP-nonreleasable Ca"~(D). Cells tides (primarily NADPH) is responsible for this decrease in were pretreated as control (open symbols) or with BCNU (closed sym- mitochondrial Ca2+ during hydroperoxide metabolism. Thus, bols), as described, prior to the addition of hydroperoxide (nm, 1 mM; the inhibition of both pyridine nucleotide oxidation and mito- 0,0e,4mM) at time zero. The results are typical of values obtained for chondrial Ca2+ release by 1,4-dithiothreitol and BCNU indi- three to five trials and are expressed as percentage of the time zero value. The absolute values for the BCNU-pretreated hepatocytes at cates that the factor responsible for the decrease in mitochon- this time per 1 x 106 cells were: GSH, 52.5 nmol; NADPH, 3.5 nmol; drial Ca2+ level is distal to glutathione reductase. Therefore, FOOP-releasable Ca"~, 1.7 nmol; FOOP-nonreleasable Ca"~, 1.0 nmol. our data confirm the role ofthe pyridine nucleotide redox state in the control ofmitochondrial Ca2+ concentration in the intact to cause Ca2 release (29). In addition, the amount of Ca2' re- cell, as predicted by Lehninger et al. (14). leased from isolated hepatocytes by FCCP as measured with Incubation of rat hepatocytes with the alkylating agent arsenazo III is very close to the mitochondrial Ca2' level de- BCNU has been shown to cause inactivation of glutathione re- termined by atomic absorption spectrometry after rapid cell ductase (28), and Eklow et aL have recently developed an in- disruption (13). The Ca21 that is releasable by both FCCP and cubation procedure that yields hepatocytes with inhibited glu- dicumarol in our system is in close agreement with these mea- tathione reductase but that exhibit a normal GSH level (24). surements. Incubation ofhepatocytes, pretreated with BCNU according to Further, our findings show that incubation with ruthenium this procedure, with t-butyl hydroperoxide resulted in en- red, which inhibits only mitochondrial Ca2' uptake (22, 30), hanced loss of cellular GSH and FCCP-nonreleasable Ca2' leads to a substantial decrease in the size of the FCCP-releas- without apparent effects on the pyridine nucleotide redox level able pool as well as a reduction in total cell Ca2', indicating that or the size ofthe FCCP-releasable Ca2+ pool (cf. Fig. 4). There- the plasma membrane pumps extrude Ca2' that is lost by the fore, it appears that oxidation of GSH, rather than pyridine mitochondria. Results obtained from the addition of digitonin nucleotide oxidation, is responsible for the release of extrami- and FCCP to ruthenium red-pretreated hepatocytes also in- tochondrial Ca2+ from the hepatocytes. A similar effect ofGSH

A B FIG. 5. Scanning electron micrographs of treated hepatocytes. Micrographs were taken 30 min after the addition of 4 mM t-butyl hydroperoxide (A; x2,100), 4 mM t-butyl hydroperoxide and 2 mM 1,4-dithiothreitol (B; x2,800), or 1 mM t-butyl hydroperoxide and 100 ,uM BCNU (C; x2,100). Downloaded by guest on September 27, 2021 6846 Biochemistry: Bellomo et al. Proc. Natl. Acad. Sci. USA 79 (1982) depletion on cellular calcium homeostasis has recently been 9. Simon, W., Ammann, D., Oehme, M. & Morf, W. E. (1978) observed with several other agents that affect the GSH status Ann. N.Y. Acad. Sci. 307, 52-69. of isolated hepatocytes (33). 10. Blinks, J. R. (1978) Ann. N.Y. Acad. Sci. 307, 71-84. 11. Kendrick, M. C., Ratzlaff, R. W. & Blaustein, M. P. (1977) Anal. Cumulative evidence supports the assumption that the per- Biochem.. 83, 433-450. turbation ofhepatocyte surface structure during t-butyl hydro- 12. Tischler, M. E., Hecht, P. & Williamson, J. R. (1977) Arch. peroxide metabolism is related to GSH depletion and loss of Biochem. Biophys. 181, 278-292. Ca2" from the extramitochondrial compartment (33). Soluble 13. Murphy, E., Coll, K., Rich, T. L. & Williamson, J. R. (1980) J. thiols, notably GSH, may exert a protective effect by (i) pre- Biol Chem. 255, 6600-6608. venting the inactivation of Ca2"-binding proteins such as cal- 14. Lehninger, A. L., Vercesi, A. & Bababunmi, E. A. (1978) Proc. Natl Acad. Sci. USA 75, 1690-1694. modulin, which has functionally important methionine groups 15. Lotscher, H. R., Winterhalter, K. H., Carafoli, E. & Richter, C. (34), and (ii) by maintaining the thiol groups of intracellular (1979) Proc. Nati Acad. Sci. USA 76, 4340-4344. membrane proteins in their normal state, most probably via the 16. Prpic, V. & Bygrave, F. L. (1980)J. Biol Chem. 255, 6193-6199. cytosolic thiol transferase (35). The latter may be particularly 17. Bellomo, G., Jewell, S. A. & Orrenius, S. (1982)J. Biol. Chem., importantfor the maintenance ofthe calcium sequestering func- in press. tion of the endoplasmic reticulum (36). The destruction of cal- 18. Sies, H., Graf, P. & Estrela, J. M. (1981) Proc. Natl. Acad. Sci. USA 78, 3358-3362. cium sequestration in the endoplasmic reticulum would cause 19. Tamura, M., Oshino, N., Chance, B. & Silver, I. A. (1978) Arch. a gross disturbance of intracellular Ca2" homeostasis and the Biochem. Biophys. 191, 8-22. inhibition ofenzymes associated with the hepatocyte cytoskele- 20. Orrenius, S., Thor, H., Ekldw, L., Moldeus, P. & Jones, D. P. tal apparatus, such as actomyosin ATPase. The link between (1982) Adv. Exp. Med. Biol 136, 395-405. intracellular thiol and Ca2" homeostasis and the functional op- 21. Moldeus, P., Hogberg, J. & Orrenius, S. (1978) Methods En- eration of the further zymol 51, 60-71. hepatocyte cytoskeletal apparatus awaits 22. Hogberg, J. & Kristoferson, A. (1977) Eur. J. Biochem. 74, 77- investigation. 82. The authors gratefully acknowledge the expert assistance provided 23. Moore, C. L. (1970) Biochem. Biophys. Res. Commun. 42, 298- by Dr. Dean P. Jones in developing the microsomal measurement 305. Ca2" 24. Eklow, L., Mold6us, P. & Orrenius, S. (1982) in Cytochrome P- system and in reviewing this manuscript. We also thank Ms. Pia Hart- 450. Biochemistry, Biophysics and Environmental Implications, zell and Dr. Janet R. Dawson for technical assistance and Mr. Sten ed. Hietanen, E. (Elsevier North-Holland, Amsterdam), in Thorold for preparation offigures and electron micrographs. This work press. was supported by grants from the Swedish Medical Research Council 25. Saville, B. (1958) Analyst 83, 670-672. (03X-2471), the Swedish Council for the Planning and Coordination of 26. Reed, D. J., Babson, J. R., Beatty, P. W., Brodie, A. E., Ellis, Research, and the Nobel Foundation. W. W. & Pollis, D. W. (1980) AnaL Biochem. 106, 55-62. 27. Klingenberg, M. (1974) in Methoden der Enzymatischen Ana- 1. Scarpa, A. & Carafoli, E. (1978) Calcium Transport and Cell lyse, ed. Bergmeyer, H. U. (Verlag Chemie, Weinheim), pp. Function (New York Academy of Science, New York). 2094-2122. 2. Carafoli, E., Clement, F., Drabikowski, W. & Margreth, A. 28. Babson, J. R., Abell, N. S. & Reed, D. J. (1981) Biochem. Phar- (1975) Calcium Transport in Contraction and Secretion (North- macol 30, 2299-2304. Holland, Amsterdam). 29. Chen, J. L., Babcock, D. F. & Lardy, H. A. (1978) Proc. Natl. 3. Duncan, C. J. (1976) Calcium in Biological Systems, 30th Sym, Acad. Sci. USA 75, 2234-2238. posium of the Society for Experimental Biology (Cambridge 30. Bygrave, F. L. (1978) Biochem.J. 170, 87-91. Univ. Press, London). 31. Kleineke, J. & Stratman, F. W. (1964) FEBS Lett. 43, 75-80. 4. Dawson, R. M. C. & Hauser, H. (1970) in Calcium and Cellular 32. Dubinsky, W. P. & Cockrell, R. S. (1975) FEBS Lett. 59, 39-43. Function, ed. Cuthbert, A. W. (Mac Millan, London), pp. 17-41. 33. Jewell, S. A., Bellomo, G., Thor, H., Orrenius, S. & Smith, M. 5. Borle, A. B. (1972) Fed. Proc. Fed. Am. Soc. Exp. Biol. 32, 1944- T. (1982) Science 217, 1257-1259. 1950. 34. Walsh, M. & Stevens, F. C. (1978) Biochemistry 17, 3924-3929. 6. Bygrave, F. L. (1978) Biol. Rev. 53, 43-79. 35. Mannervik, B., Axelsson, K. & Larsson, K. (1981) Methods En- 7. Becker, G. L., Fiskum, G. & Lehninger, A. L. (1980) J. Biol zymol 77, 281-285. Chem. 255, 9009-9012. H. & E. 8. Claret-Berthon, B., Claret, M. & Mazet, J. L. (1977)J. Physiol 36. Moore, L., Chen, T., Knapp, R., Jr., Landon, J. (1975) (London) 272, 529-552. J. Biol Chem. 250, 4562-4568. Downloaded by guest on September 27, 2021